A DNA Barcode Library for North American Ephemeroptera: Progress and Prospects Jeffrey M. Webb1*, Luke M. Jacobus2, David H. Funk3, Xin Zhou4, Boris Kondratieff5, Christy J. Geraci6,R. Edward DeWalt7, Donald J. Baird8, Barton Richard9, Iain Phillips10, Paul D. N. Hebert1 1 Biodiversity Institute of Ontario, University of Guelph, Guelph, Ontario, Canada, 2 Division of Science, Indiana University Purdue University Columbus, Columbus, Indiana, United States of America, 3 Stroud Water Research Center, Avondale, Pennsylvania, United States of America, 4 BGI, Shenzhen, Guangdong Province, China, 5 Department of Bioagricultural Sciences and Pest Management, Colorado State University, Fort Collins, Colorado, United States of America, 6 Department of Entomology, National Museum of Natural History, Smithsonian Institution, Washington, D. C., United States of America, 7 Prairie Research Institute, Illinois Natural History Survey, University of Illinois, Champaign, Illinois, United States of America, 8 Environment Canada, Canadian Rivers Institute, Department of Biology, University of New Brunswick, Fredericton, New Brunswick, Canada, 9 Laboratory of Aquatic Entomology, Florida A&M University, Tallahassee, Florida, United States of America, 10 Saskatchewan Watershed Authority, Saskatoon, Saskatchewan, Canada Abstract DNA barcoding of aquatic macroinvertebrates holds much promise as a tool for taxonomic research and for providing the reliable identifications needed for water quality assessment programs. A prerequisite for identification using barcodes is a reliable reference library. We gathered 4165 sequences from the barcode region of the mitochondrial cytochrome c oxidase subunit I gene representing 264 nominal and 90 provisional species of mayflies (Insecta: Ephemeroptera) from Canada, Mexico, and the United States. No species shared barcode sequences and all can be identified with barcodes with the possible exception of some Caenis. Minimum interspecific distances ranged from 0.3–24.7% (mean: 12.5%), while the average intraspecific divergence was 1.97%. The latter value was inflated by the presence of very high divergences in some taxa. In fact, nearly 20% of the species included two or three haplotype clusters showing greater than 5.0% sequence divergence and some values are as high as 26.7%. Many of the species with high divergences are polyphyletic and likely represent species complexes. Indeed, many of these polyphyletic species have numerous synonyms and individuals in some barcode clusters show morphological attributes characteristic of the synonymized species. In light of our findings, it is imperative that type or topotype specimens be sequenced to correctly associate barcode clusters with morphological species concepts and to determine the status of currently synonymized species. Citation: Webb JM, Jacobus LM, Funk DH, Zhou X, Kondratieff B, et al. (2012) A DNA Barcode Library for North American Ephemeroptera: Progress and Prospects. PLoS ONE 7(5): e38063. doi:10.1371/journal.pone.0038063 Editor: Brock Fenton, University of Western Ontario, Canada Received February 2, 2012; Accepted April 30, 2012; Published May 30, 2012 Copyright: ß 2012 Webb et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This project was funded by the Government of Canada through Genome Canada and the Ontario Genomics Institute (2008-OGI-ICI-03), and Environment Canada’s Competitiveness and Environmental Sustainability Indicators (CESI) Program. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected] Introduction addition to allowing the identification of difficult specimens, barcoding provides a level of data standardization that has been DNA barcoding [1] of animals, the analysis of a standardized previously lacking in environmental assessments [13], aiding segment of the mitochondrial cytochrome c oxidase subunit 1 broader comparisons of results gathered through monitoring (COI) gene, has rapidly become an important tool for the programs. Furthermore, barcoding reliably produces species-level identification, delimitation, and discovery of species [1–3]. DNA (or even population-level) identifications that can improve the barcoding has the further advantage that identification success sensitivity of analyses, depending on the aims of the assessment extends across all life stages, allowing the association of immatures program [14,15]. with adults [4,5]. Its capacity to identify all life stages is particularly An essential requirement of biomonitoring, conservation bi- important for aquatic ecology and biological monitoring (biomo- ology, and ecology is a sound taxonomy of the study organisms. nitoring) of water quality because the aquatic larvae are usually the Although taxa could be based purely on operational taxonomic life stage studied [6]. Unfortunately, this is the life stage that is the units (OTUs) defined by sequence data, information on key most poorly known taxonomically because most species concepts functional niche traits (e.g. [16]; functional feeding groups, life in aquatic insects are based on the morphology of adult males. The history data, behavior and historical distribution) and environ- identification of larvae is further hindered by the fact that many mental sensitivity [17] requires linking molecular OTUs with are rather delicate, especially mayflies (Ephemeroptera), and the known taxa. structures critical for confident identifications such as gills, legs, Mayflies are abundant in most aquatic habits, and show varied and caudal filaments are commonly damaged or missing. tolerance to differing disturbance regimes, making them ideal The application of DNA barcoding to freshwater biomonitoring candidates for monitoring water quality. Together with the has recently generated much interest for several reasons [7–12]. In caddisflies (Trichoptera) and stoneflies (Plecoptera), they comprise PLoS ONE | www.plosone.org 1 May 2012 | Volume 7 | Issue 5 | e38063 DNA Barcoding North American Ephemeroptera one of the most commonly used biomonitoring metrics, EPT intraspecific divergences greater than 2.2%, a level of divergence richness [18], and as such are one of the focal groups for the found to delimit species across diverse groups of insects [1,2,20] International Barcode of Life Freshwater Biosurveillance Working although higher maximum intraspecific distances have been Group (www.ibol.org). observed in Trichoptera when widely separated geographic areas The development of a barcode reference library for North are included [3]. Almost 20% of species with more than one American mayflies was initiated by sequencing 150 specimens specimen had maximum intraspecific divergences .5.0%; these from 80 North American species [2]. Those results indicated that species with high intraspecific divergences may represent species- all species tested could be identified accurately using barcodes, but complexes and when they are excluded from the analysis, the revealed several cases of deep sequence divergence within a species. mean maximum intraspecific divergence decreased to 1.3%. In 44 Further species were added [19] and barcoding techniques used as species (12.5%), the maximum intraspecific distance was greater a taxonomic tool to confirm the validity of a presumed new species than the minimum interspecific distance. A Neighbor Joining tree and to support the synonymy of two species of Heptagenia of all specimens is available in Figure S1. (Heptageniidae). Additional barcodes for North American mayflies Increasing the geographic range between samples did not were generated from a regional inventory of northeastern always lead to large increases in intraspecific divergence. For Manitoba [20,21], from a test of the efficacy of barcoding for example, Epeorus vitreus (Walker) was sampled from throughout biological monitoring using aquatic macroinvertebrates [12], and its latitudinal range but the maximum intraspecific divergence a general barcoding paper [1]. Additional barcode sequences for was only 2.8% and a specimen of Ephemera simulans Walker from North American species of Ephemerella have been generated [22], Colorado differed by only a single nucleotide (0.15%) from but because these sequences are not publicly available, we do not a specimen from Churchill, Manitoba, a distance of over consider them further. 2,200 km. Even in species with multiple barcode clusters, The mayfly fauna of North America includes 651species geographically distant specimens often clustered within the (Mayfly Central, http://www.entm.purdue.edu/mayfly/na- same group and specimens with small geographic distances species-list.php accessed 30 November 2011), but 10 of these taxa sometimes occurred in different barcode clusters. The cluster of are nomina dubia and four are recently extinct. Ignoring the latter Acentrella parvula (McDunnough) that included a topotypical two groups, 637 species and 8 subspecies remain. Previous papers (from type locality) specimen from southern Ontario, for have assembled barcodes for 121 of these species, most from example, also included specimens from Saskatchewan and eastern North America [1,2,12,19,20]. In this paper, we broaden New Brunswick, but other specimens